Table 3. Cross-Lagged Models: Fit Indices and Model Comparisons.
| Model fit indices | Model comparison | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Models | χSB2 | df | TLI | CFI | RMSEA [90% CI] | Models | ΔχSB2 | Δdf | p | ΔCFI | ΔRMSEA |
| Empathy, balanced relatedness, antisocial behaviors (A) | |||||||||||
| M1: Unconstrained | 203.549 | 78 | .934 | .966 | .057 [.047, .067] | ||||||
| M2: Constrained | 252.832 | 114 | .950 | .962 | .050 [.041, .058] | M2-M1 | 49.65 | 36 | .065 | -.004 | -.007 |
| Empathy, balanced relatedness, antisocial behaviors (M) | |||||||||||
| M1: Unconstrained | 270.151 | 78 | .916 | .956 | .070 [.061, .080] | ||||||
| M2: Constrained | 321.797 | 114 | .938 | .953 | .061 [.053, .068] | M2-M1 | 51.178 | 36 | .048 | -.003 | -.009 |
| Empathy, conflict, antisocial behaviors (A) | |||||||||||
| M1: Unconstrained | 183.206 | 78 | .951 | .974 | .052 [.042, .062] | ||||||
| M2: Constrained | 243.469 | 114 | .958 | .968 | .048 [.039, .056] | M2-M1 | 60.622 | 36 | .006 | -.006 | -.004 |
| Empathy, conflict, antisocial behaviors (M) | |||||||||||
| M1: Unconstrained | 245.237 | 78 | .935 | .966 | .066 [.057, .075] | ||||||
| M2: Constrained | 286.841 | 114 | .954 | .965 | .055 [.047, .063] | M2-M1 | 47.220 | 36 | .100 | -.001 | -.011 |
| Empathy, support, antisocial behaviors (A) | |||||||||||
| M1: Unconstrained | 176.224 | 78 | .953 | .976 | .050 [.040, .060] | ||||||
| M2: Constrained | 230.516 | 114 | .962 | .971 | .045 [.037, .054] | M2-M1 | 55.793 | 36 | .019 | -.005 | -.005 |
| Empathy, support, antisocial behaviors (M) | |||||||||||
| M1: Unconstrained | 242.510 | 78 | .937 | .967 | .065 [.056, .075] | ||||||
| M2: Constrained | 275.526 | 114 | .958 | .968 | .053 [.045, .061] | M2-M1 | 36.964 | 36 | .424 | .001 | -.012 |
Note. A = adolescent report; M = maternal report. χ2 = Chi-Square; df = degrees of freedom; TLI = Tucker-Lewis Index; CFI = Comparative Fit Index; RMSEA = Root Mean Square Error of Approximation and 90% Confidence Interval; Δ = change in parameter. ΔχSB2 model comparisons are based on Satorra and Bentler’s (2001) scaled difference chi-square test statistic.